164 
JOURNAL OF MAMMALOGY 
acter should not bear much weight in an animal as variable in minor 
details as the meadow mouse. 
COLORATION 
The color of the meadow mouse is due to a mixture of yellowish 
rusty and blackish hairs, the relative abundance of which determines 
the particular shade. In a large series of meadow mice there is a great 
variety of fine gradations of color from very pale, yellowish specimens 
to those which are dark brown or blackish. Many from the mainland 
of New England are almost black in color. The series of M. p. shat- 
tucki, when compared with a series of three hundred skins from a great 
variety of localities, were seen to be in about the middle of the range 
of color variation. They are rather bright with a large proportion 
of yellow and reddish hairs, being almost identical in color with speci- 
mens from various localities in Massachusetts, and distinctly lighter 
than many others also from Massachusetts. 
AUDITAL BULL^ 
Upon superficial examination there appears to be no noticeable 
difference between the bullae of the skulls from Tumble Down Dick and 
of those collected in various places in New England. A skull from 
Wareham, Massachusetts, was seen to be almost an exact match in 
every particular for the skull of the type specimen of M. p, shattucki. 
The skulls of the series from Grand Manan are comparatively large 
and heavy and the bullae are large in proportion, appearing to be more 
globular than those of shattucki. For purposes of comparison the great- 
est length and the greatest width of the right bullae of the seven skulls 
of shattucki and of ninety-six skulls of pennsylvanicus were measured 
to the nearest hundredth of a millimeter by means of calipers and a 
micrometer. The dimensions of the bullae of shattucki ranged from 
7.32 by 5.75 to 8.85 by 6.63 mm. Those of pennsylvanicus run up to 
9.41 by 7.00 mm. The average dimensions of the right bullae of fifteen 
specimens from Grand Manan is 8.53 by 6.43 mm. In order to obtain 
a quantitative expression of the globularity of the bullae, the width 
was subtracted from the length in each case, and the result was divided 
by the length. This gave an index of the ovalness of the bulla which 
would be smaller in the case of more globular and larger in the case 
of less globular bullae. The indices thus obtained were grouped into 
nine magnitude classes and the classes were numbered in order. Class 
